Questions about patterns

[Stikus]

Hello again, looks like you have some patterns disabled for now (commented out in script):
https://github.com/nickp60/fcid/blob/main/fcid/run.py#L115-L122

    #    ["[A-Z,0-9]{5}ACXX",   [["HiSeq 1000", "HiSeq 1500", "HiSeq 2000", "HiSeq 2500"], "High Output (8-lane) v3 flow cell"]],
    #    ["H[A-Z,0-9]{4}BCXY",   [["HiSeq 1500", "HiSeq 2500"], "Rapid Run (2-lane) v2 flow cell"]],
    #    ["C[A-Z,0-9]{4}ANXX",   [["HiSeq 1500", "HiSeq 2000", "HiSeq 2500"], "High Output (8-lane) v4 flow cell"]],
    #    ["C[A-Z,0-9]{4}AC[A-Z,0-9]{2}",   [["HiSeq 2500"], "High Output (8-lane) v4 flow cell"]],
    #    ["H[A-Z,0-9]{4}ADXX",   [["HiSeq 1500", "HiSeq 2500"], "Rapid Run (2-lane) v1 flow cell"]],
    #    ["H[A-Z,0-9]{4}ADXY",   [["HiSeq 1500", "HiSeq 2500"], "Rapid Run (2-lane) v1 flow cell"]],
    #    ["H[A-Z,0-9]{4}BC[A-Z,0-9]{2}",   [["HiSeq 2500"], "Rapid Run (2-lane) v2 flow cell"]],
    #    ["H[A-Z,0-9]{4}BBXY[A-Z,0-9]{2}",   [["HiSeq 4000"], "(8-lane) v1 flow cell"]],

What is the reason behind it? Some support emails? Can you share them for clarity?

For example, we have public data from GIAB (Garvan_NA12878_HG001_HiSeq_Exome) with fastq header:

@HWI-D00119:50:H7AP8ADXX:1:1101:1213:2058 1:N:0:TAAGGCGA

For now, it cannot be detected with flowcell id due to commented pattern, but can be detected with by-machine:

fcid H7AP8ADXX
Unknown Machine
fcid HWI-D00119 --by-machine
HiSeq 2500

And one more public data from GIAB (NIST_NA12878_HG001_HiSeq_300x) - it cannot be detected at all:

@HISEQ1:17:H947YADXX:1:1101:1442:2106 1:N:0:CGATGT

fcid HISEQ1 --by-machine
Unknown
fcid H947YADXX
Unknown Machine

---

And one more question - what do you think about detection non-Illumina sequencers? MGI/BGI or ONT for example?

[Stikus]

Btw, some minor problems:

• https://github.com/nickp60/fcid/blob/main/fcid/run.py#L24

    [".*", [["Unknown"]]],

• https://github.com/nickp60/fcid/blob/main/fcid/run.py#L123

    [".*", [["Unknown Machine"], "Unknown flow cell"]],

Shouldn't Unknown be changed to Unknown Machine for consistency?

---

• https://github.com/nickp60/fcid/blob/main/fcid/run.py#L22-L23

    ["SN[0-9]{3}", [["HiSeq2000", "HiSeq2500"]]],
    ["SN[0-9]{3}", [["HiSeq2000", "HiSeq2500"]]],

Shouldn't it be "HiSeq 2000", "HiSeq 2500" (with spaces)? And these strings are same.

---

https://github.com/nickp60/fcid/blob/main/fcid/run.py#L20-L21

    ["A[0-9]{5}", [["NovaSeq"]]],
    ["NA[0-9]{5}", [["NovaSeq"]]],

Shouldn't it be `"NovaSeq 6000"? (according to OpenGene/fastp#508 (comment) for example)

---

https://github.com/nickp60/fcid/blob/main/fcid/run.py#L16-L17

    ["NB[0-9]{6}", [["NextSeq"]]],
    ["NS[0-9]{6}", [["NextSeq"]]],

Shouldn't it be "NextSeq 500", "NextSeq 550"? (according to OpenGene/fastp#508 (comment) for example)

---

https://github.com/nickp60/fcid/blob/main/fcid/run.py#L98-L99

    ["H[A-Z,0-9]{4}BGXX", [["NextSeq"], "High output flow cell"]],
    ["H[A-Z,0-9]{4}BGXY", [["NextSeq"], "High output flow cell"]],

These patterns should be removed because it is part of

    [
        "[A-Z,0-9]{5}BG[A-Z,0-9]{2}",
        [["NextSeq 500", "NextSeq 550"], "High Output flow cell"],
    ],

[Stikus]

• https://github.com/nickp60/fcid/blob/main/fcid/run.py#L67-L74

    ["BNT[A-Z,0-9]{5}-[A-Z,0-9]{4}", [["iSeq 100"], "Standard Output flow cell"]],
    ["BRB[A-Z,0-9]{5}-[A-Z,0-9]{4}", [["iSeq 100"], "Standard Output flow cell"]],
    ["BPC[A-Z,0-9]{5}-[A-Z,0-9]{4}", [["iSeq 100"], "Standard Output flow cell"]],
    ["BPG[A-Z,0-9]{5}-[A-Z,0-9]{4}", [["iSeq 100"], "Standard Output flow cell"]],
    ["BPA[A-Z,0-9]{5}-[A-Z,0-9]{4}", [["iSeq 100"], "Standard Output flow cell"]],
    ["BPL[A-Z,0-9]{5}-[A-Z,0-9]{4}", [["iSeq 100"], "Standard Output flow cell"]],
    ["BNT[A-Z,0-9]{5}-[A-Z,0-9]{4}", [["iSeq 100"], "Standard Output flow cell"]],
    ["BTR[A-Z,0-9]{5}-[A-Z,0-9]{4}", [["iSeq 100"], "Standard Output flow cell"]],

There are 2 same BNT patterns.

---

Are you sure that you want , in your patterns? I'm not a very experienced python user, but it should match literal , as far as I know.

---

• https://github.com/nickp60/fcid/blob/main/fcid/run.py#L79-L83

    ["B[A-Z,0-9]{4}", [["MiSeq"], "MiSeq Standard flow cell"]],
    ["C[A-Z,0-9]{4}", [["MiSeq"], "MiSeq Standard flow cell"]],
    ["J[A-Z,0-9]{4}", [["MiSeq"], "MiSeq Standard flow cell"]],
    ["K[A-Z,0-9]{4}", [["MiSeq"], "MiSeq Standard flow cell"]],
    ["L[A-Z,0-9]{4}", [["MiSeq"], "MiSeq Standard flow cell"]],

These patterns can be combined to ["[BCJKL][A-Z,0-9]{4}", [["MiSeq"], "MiSeq Standard flow cell"]],

---

• https://github.com/nickp60/fcid/blob/main/fcid/run.py#L101

"TrueSeq v3 flow cell" - TruSeq, not TrueSeq

[nickp60]

Hi @Stikus ! Thanks for the notes; could you combine them into a PR? All would be helpful. The AD commented out pattern was a weird one -- the Illumina folks couldn't confirm it so I commented it out but I'm happy to have the re-included.

I just spent 45 minutes chatting with ONT support -- they do not have any naming conventions unfortunately, but they can look up flowcells internally against their database. I added a note about this to the readme. Not sure about MGI/BGI but if you know they follow some convention I'm happy for it to be included here! great questions.

[Stikus]

Here is PR: #7 - if all ok - can you please release new version on PyPI too (https://pypi.org/project/fcid/)?

And about ONT - we've found this documentation: https://nanoporetech.com/document/Guppy-protocol - no sequencer of flowcell-like naming, unfortunately

About DNBSEQ: https://www.seqanswers.com/forum/sequencing-technologies-companies/mgiseq-fka-complete-genomics/326115-g400-fastq-header-description

We are using your tool for autodetection instruments/flowcells from FASTQ header - for us, it is useful if there are any naming patterns in header too, but for this tool it is not (at least for now)

[Stikus]

@nickp60 Hello, can you check my MR please?

[nickp60]

@Stikus would you mind sharing your email? I'm trying to list you as a maintainer in the pyproject.toml, then I can get the new release to build. I'm not sure what to make about the BGI/MG sequencers -- the header does look interpretable but that post says that the company says there isn't interpretable metadata in the header?

For ONT, the info we would be interested in is in the sequencing_summary.txt output file -- the ONT folks said they can look up the flowcell ID from the fastq header internally to know what flowcell type and sequencing chemistry it was generated with, but the only user-interpretable output with that metadata is the summary file

[Stikus]

Thx for merge, my email is stikus@mail.ru.

We are trying to get more ONT or DNBSEQ data in a future - maybe we can provide some info about headers when we'll get it.

You can close this issue if you want - I'll open new if we'll have some new problems.

[nickp60]

Sounds good -- I just moved those to two new issues. Updated version is now on PyPi. thanks for your contributions!